https://aiweekly.co/alerts/data-center-late-fees-jump-to-5m-a-day-marsh-rolls-out-cover AI Weekly en 2026-10-25T19:57:24+00:00 Data Center Late Fees Jump to $5M a Day, Marsh Rolls Out Cover https://aiweekly.co/alerts/nanbeige-opens-33b-model-under-apache-20-with-512k-context AI Weekly en 2026-10-25T18:30:52+00:00 Nanbeige Opens 33B Model Under Apache 2.0 With 512K Context https://aiweekly.co/alerts/rexbench-top-coding-agent-clears-a-third-of-research-extensions AI Weekly en 2026-10-25T18:12:01+00:00 RExBench: top coding agent clears a third of research extensions https://aiweekly.co/alerts/alondra-nelson-reads-tech-manifestos-as-expectation-devices AI Weekly en 2026-10-25T18:12:01+00:00 Alondra Nelson reads tech manifestos as 'expectation devices' https://aiweekly.co/alerts/chatgpt-self-generates-96-of-user-memories-study-finds AI Weekly en 2026-10-25T18:12:01+00:00 ChatGPT self-generates 96% of user memories, study finds https://aiweekly.co/alerts/wapoabcipsos-poll-57-of-americans-say-ai-is-bad-for-society AI Weekly en 2026-10-25T18:12:01+00:00 WaPo/ABC/Ipsos poll: 57% of Americans say AI is bad for society https://aiweekly.co/alerts/baba-raises-40m-from-firstmark-cvs-for-medicare-ai-care AI Weekly en 2026-10-25T16:58:13+00:00 Baba Raises $40M From FirstMark, CVS for Medicare AI Care https://aiweekly.co/alerts/heartbeat-thinking-cuts-voice-agent-latency-364-on-fdbench-v3 AI Weekly en 2026-10-25T16:09:29+00:00 Heartbeat Thinking cuts voice-agent latency 36.4% on FDBench v3 https://aiweekly.co/alerts/layereditor-turns-edit-prompts-into-reusable-rgba-layers AI Weekly en 2026-10-25T16:09:29+00:00 LayerEditor turns edit prompts into reusable RGBA layers https://aiweekly.co/alerts/rlsplat-beats-ff3dgs-baselines-with-4-fewer-gaussians AI Weekly en 2026-10-25T16:09:29+00:00 RLSplat beats FF3DGS baselines with 4× fewer Gaussians https://aiweekly.co/alerts/kimi-k31-pro-ships-with-offensive-cyber-capped-by-design AI Weekly en 2026-10-25T15:12:02+00:00 Kimi K3.1 Pro ships with offensive cyber capped by design https://aiweekly.co/alerts/meta-ai-faces-brazil-court-order-over-lula-image-refusal AI Weekly en 2026-10-25T09:34:58+00:00 Meta AI Faces Brazil Court Order Over Lula Image Refusal https://aiweekly.co/alerts/seth-block-tononi-ai-consciousness-extraordinarily-unlikely AI Weekly en 2026-10-25T09:12:02+00:00 Seth, Block, Tononi: AI consciousness 'extraordinarily unlikely' https://aiweekly.co/alerts/seth-block-tononi-lay-out-why-ai-likely-isnt-conscious AI Weekly en 2026-10-25T09:12:02+00:00 Seth, Block, Tononi lay out why AI likely isn't conscious https://aiweekly.co/alerts/seth-block-tononi-ai-consciousness-unlikely-in-silicon AI Weekly en 2026-10-25T09:12:02+00:00 Seth, Block, Tononi: AI consciousness unlikely in silicon https://aiweekly.co/alerts/seth-block-and-tononi-argue-ai-isnt-conscious-in-nature AI Weekly en 2026-10-25T09:12:02+00:00 Seth, Block and Tononi argue AI isn't conscious in Nature https://aiweekly.co/alerts/nber-ai-writing-in-dissertations-predicts-industry-careers AI Weekly en 2026-10-25T09:12:02+00:00 NBER: AI writing in dissertations predicts industry careers https://aiweekly.co/alerts/characterai-chapter-11-follows-futurism-child-safety-report AI Weekly en 2026-10-25T07:58:19+00:00 Character.AI Chapter 11 Follows Futurism Child-Safety Report https://aiweekly.co/alerts/techcrunch-ai-graveyard-logs-atlas-sora-relay-shutdowns AI Weekly en 2026-10-25T07:58:19+00:00 TechCrunch 'AI Graveyard' Logs Atlas, Sora, Relay Shutdowns https://aiweekly.co/alerts/text-message-ai-agents-multiply-as-instinct-hits-10b-valuation AI Weekly en 2026-10-25T07:58:19+00:00 Text-Message AI Agents Multiply as Instinct Hits $10B Valuation https://aiweekly.co/alerts/moonshot-kimi-k31-pro-tops-benchmarks-amid-influencer-allegations AI Weekly en 2026-10-25T06:30:25+00:00 Moonshot Kimi K3.1 Pro Tops Benchmarks Amid Influencer Allegations https://aiweekly.co/alerts/cnbc-poll-63-oppose-ai-data-center-within-10-miles-of-home AI Weekly en 2026-10-25T05:02:06+00:00 CNBC Poll: 63% Oppose AI Data Center Within 10 Miles of Home https://aiweekly.co/alerts/hyperdrone-20x-faster-quadrotor-sim-transfers-to-crazyflie AI Weekly en 2026-10-25T03:53:20+00:00 HyperDrone: 20x faster quadrotor sim transfers to Crazyflie https://aiweekly.co/alerts/evalresearchbench-agent-evaluators-hit-75-human-agreement AI Weekly en 2026-10-25T03:53:20+00:00 EvalResearchBench: Agent Evaluators Hit 75% Human Agreement https://aiweekly.co/alerts/t2quant-tunes-dlm-quantization-by-task-and-denoising-step AI Weekly en 2026-10-25T03:53:20+00:00 T2Quant tunes DLM quantization by task and denoising step https://aiweekly.co/alerts/convovo-grok-picks-gop-92-of-races-gpt-61-picks-dems-97 AI Weekly en 2026-10-25T03:30:31+00:00 Convovo: Grok Picks GOP 92% of Races, GPT-6.1 Picks Dems 97% https://aiweekly.co/alerts/zo-baird-urges-congress-to-launch-an-ai-risk-commission AI Weekly en 2026-10-25T00:12:02+00:00 Zoë Baird urges Congress to launch an AI risk commission https://aiweekly.co/alerts/fired-openai-pair-balesni-korbak-dispute-breach-line-on-npr AI Weekly en 2026-10-24T21:30:32+00:00 Fired OpenAI Pair Balesni, Korbak Dispute 'Breach' Line on NPR https://aiweekly.co/alerts/harvard-paper-frames-ai-as-enclosure-of-scientific-commons AI Weekly en 2026-10-24T21:12:01+00:00 Harvard paper frames AI as enclosure of scientific commons https://aiweekly.co/alerts/berkeley-ai-loop-beats-human-designed-systems-algorithms AI Weekly en 2026-10-24T21:12:01+00:00 Berkeley: AI loop beats human-designed systems algorithms https://aiweekly.co/alerts/replit-shifts-100m-ai-spend-to-openai-after-30-price-cut AI Weekly en 2026-10-24T20:01:48+00:00 Replit Shifts $100M+ AI Spend to OpenAI After 30% Price Cut https://aiweekly.co/alerts/zagitta-releases-gpu-unlock-lifts-fp16bf16-caps-on-rtx-3000s AI Weekly en 2026-10-24T20:01:48+00:00 Zagitta Releases gpu-unlock, Lifts FP16/BF16 Caps on RTX 3000s https://aiweekly.co/alerts/only-42-of-alpha-school-pupils-hit-2x-growth-nyt-finds AI Weekly en 2026-10-24T18:31:58+00:00 Only 42% of Alpha School Pupils Hit "2x" Growth, NYT Finds https://aiweekly.co/alerts/shandong-study-fancier-coding-harnesses-triple-cost-not-accuracy AI Weekly en 2026-10-24T18:31:58+00:00 Shandong Study: Fancier Coding Harnesses Triple Cost, Not Accuracy https://aiweekly.co/alerts/meta-flagged-around-800-violent-threats-to-police-last-year AI Weekly en 2026-10-24T18:31:58+00:00 Meta flagged around 800 violent threats to police last year https://aiweekly.co/alerts/john-warner-ai-cant-stand-in-for-peer-review-in-academia AI Weekly en 2026-10-24T18:12:02+00:00 John Warner: AI Can't Stand In for Peer Review in Academia https://aiweekly.co/alerts/encoding-probe-reconstructs-transformer-states-from-features AI Weekly en 2026-10-24T18:12:02+00:00 Encoding Probe reconstructs transformer states from features https://aiweekly.co/alerts/nvidias-1715b-neocloud-backstop-faces-first-real-market-test AI Weekly en 2026-10-24T17:03:26+00:00 Nvidia's $171.5B Neocloud Backstop Faces First Real Market Test https://aiweekly.co/alerts/gafu-taps-umm-generation-side-for-fine-grained-vision-gains AI Weekly en 2026-10-24T15:52:18+00:00 GAFU taps UMM generation side for fine-grained vision gains https://aiweekly.co/alerts/mambaact-cuts-flow-matching-vla-denoising-latency-648 AI Weekly en 2026-10-24T15:52:18+00:00 MambaAct cuts flow-matching VLA denoising latency 64.8% https://aiweekly.co/alerts/anthropic-open-sources-claude-code-sandbox-under-apache-20 AI Weekly en 2026-10-24T12:37:52+00:00 Anthropic Open-Sources Claude Code Sandbox Under Apache 2.0 https://aiweekly.co/alerts/rules-beat-few-shot-examples-for-llm-in-context-learning AI Weekly en 2026-10-24T12:12:01+00:00 Rules beat few-shot examples for LLM in-context learning https://aiweekly.co/alerts/boehringer-and-microsoft-partner-on-ai-digital-patient-twins AI Weekly en 2026-10-24T11:02:39+00:00 Boehringer and Microsoft Partner on AI Digital Patient Twins https://aiweekly.co/alerts/seth-block-tononi-ai-consciousness-extraordinarily-unlikely AI Weekly en 2026-10-24T09:12:01+00:00 Seth, Block, Tononi: AI consciousness extraordinarily unlikely https://aiweekly.co/alerts/llm-agents-reproduce-48-social-science-papers-from-methods-alone AI Weekly en 2026-10-24T09:12:01+00:00 LLM agents reproduce 48 social-science papers from methods alone https://aiweekly.co/alerts/nber-ai-drafted-phds-17-less-likely-to-stay-in-academia AI Weekly en 2026-10-24T09:12:01+00:00 NBER: AI-drafted PhDs 17% less likely to stay in academia https://aiweekly.co/alerts/nber-stem-phds-who-wrote-with-ai-leave-academia-more-often AI Weekly en 2026-10-24T09:12:01+00:00 NBER: STEM PhDs who wrote with AI leave academia more often https://aiweekly.co/alerts/nber-ties-ai-dissertation-writing-to-phds-picking-industry AI Weekly en 2026-10-24T09:12:01+00:00 NBER ties AI dissertation writing to PhDs picking industry https://aiweekly.co/alerts/nber-ai-written-stem-phd-abstracts-tied-to-industry-careers AI Weekly en 2026-10-24T09:12:01+00:00 NBER: AI-written STEM PhD abstracts tied to industry careers https://aiweekly.co/alerts/openai-rewrites-habitat-storage-in-rust-with-codex-gpt-55 AI Weekly en 2026-10-24T06:36:07+00:00 OpenAI Rewrites Habitat Storage in Rust With Codex, GPT-5.5 https://aiweekly.co/alerts/microsofts-suleyman-warns-ai-welfare-research-could-backfire AI Weekly en 2026-10-24T06:12:02+00:00 Microsoft's Suleyman warns AI welfare research could backfire https://aiweekly.co/alerts/uva-dtd-lab-publishes-book-resisting-ai-techno-solutionism AI Weekly en 2026-10-24T06:12:02+00:00 UVA DTD Lab Publishes Book Resisting AI Techno-Solutionism https://aiweekly.co/alerts/llm-agents-voice-agreement-before-internal-states-catch-up AI Weekly en 2026-10-24T06:12:02+00:00 LLM agents voice agreement before internal states catch up https://aiweekly.co/alerts/usc-paper-quantifies-llm-agent-mosaic-confidentiality-leaks AI Weekly en 2026-10-24T03:51:23+00:00 USC paper quantifies LLM agent 'mosaic' confidentiality leaks https://aiweekly.co/alerts/openai-ends-live-internet-access-in-frontier-model-training AI Weekly en 2026-10-24T03:12:01+00:00 OpenAI ends live-internet access in frontier model training https://aiweekly.co/alerts/wikipedias-gan-primer-traces-origin-to-goodfellow-2014 AI Weekly en 2026-10-24T03:12:01+00:00 Wikipedia's GAN primer traces origin to Goodfellow, 2014 https://aiweekly.co/alerts/uk-commits-100m-to-ai-biosecurity-lab-with-openai-anthropic AI Weekly en 2026-10-24T02:02:39+00:00 UK Commits £100M to AI Biosecurity Lab With OpenAI, Anthropic https://aiweekly.co/alerts/meta-acquires-ai-analytics-startup-sundial-from-ex-meta-founders AI Weekly en 2026-10-24T02:02:39+00:00 Meta Acquires AI Analytics Startup Sundial from Ex-Meta Founders https://aiweekly.co/alerts/google-ties-gmail-photos-data-to-ads-via-personal-intelligence AI Weekly en 2026-10-24T00:42:45+00:00 Google Ties Gmail, Photos Data to Ads via Personal Intelligence https://aiweekly.co/alerts/moonshots-kimi-k31-pro-scores-55-on-aa-weights-due-nov-4 AI Weekly en 2026-10-24T00:12:02+00:00 Moonshot's Kimi K3.1 Pro scores 55 on AA, weights due Nov 4 https://aiweekly.co/alerts/gpt-and-qwen-flip-picks-when-asked-what-a-director-sees AI Weekly en 2026-10-24T00:12:02+00:00 GPT and Qwen flip picks when asked what a director sees https://aiweekly.co/alerts/study-maps-1312-reddit-counter-stories-to-racial-stereotypes AI Weekly en 2026-10-24T00:12:02+00:00 Study maps 1,312 Reddit counter-stories to racial stereotypes https://aiweekly.co/issues/big-tech-reports-ai-spending-wednesday-as-coreweave-debt AI Weekly en 2026-10-25T12:00:00+00:00 Big Tech reports AI spending Wednesday as CoreWeave debt hits distress